The Agent Skills Directory

[Prompt Injection] (SAFE): Strong instructional language like 'MANDATORY' is used to enforce architectural patterns and best practices rather than attempting to bypass safety filters or override system constraints.
[Data Exposure & Exfiltration] (SAFE): The skill uses local SQLite storage (app.db). No network requests, hardcoded credentials, or sensitive path access were detected.
[Indirect Prompt Injection] (LOW): While the skill defines functions that process user data (database queries), it provides mandatory guidance on boundary markers (parameterized queries) to prevent injection attacks.
Ingestion points: db_query, db_execute, and db_insert in database.py.
Boundary markers: Explicitly mandates use of ? placeholders and tuple parameters.
Capability inventory: File system access for SQLite database creation/modification.
Sanitization: Relies on the standard library sqlite3 parameterization mechanism.
[Unverifiable Dependencies] (SAFE): The skill relies solely on the Python standard library (sqlite3, typing) and does not attempt to install external packages or execute remote scripts.

database-management